NBER WORKING PAPER SERIES
WHEN THE LEVEE BREAKS:
BLACK MIGRATION AND ECONOMIC DEVELOPMENT IN THE AMERICAN SOUTH
Richard Hornbeck
Suresh Naidu
Working Paper 18296
http://www.nber.org/papers/w18296
NATIONAL BUREAU OF ECONOMIC RESEARCH
1050 Massachusetts Avenue
Cambridge, MA 02138
August 2012
For comments and suggestions, we thank Ran Abramitzky, Daron Acemoglu, Lee Alston, Hoyt Bleakley,
Leah Boustan, Bill Collins, Paul David, Dave Donaldson, Price Fishback, Claudia Goldin, Larry Katz,
Paul Rhode, Andrei Shleifer, Jeff Vincent, Gavin Wright, and seminar participants at AEA, Arizona,
BREAD, Columbia, CIFAR, Pittsburgh, Harvard, NBER, Stanford, UC Berkeley, and UC San Diego.
For financial support, we thank CIFAR and the WCFIA's Project on Justice, Welfare, and Economics.
Tom Beckford, James Feigenbaum, Lillian Fine, Andrew Das Sarma, and Leo Schwartz provided
excellent research assistance. The views expressed herein are those of the authors and do not necessarily
reflect the views of the National Bureau of Economic Research. For financial support, we thank CIFAR
and the WCFIA's Project on Justice, Welfare, and Economics.
NBER working papers are circulated for discussion and comment purposes. They have not been peerreviewed or been subject to the review by the NBER Board of Directors that accompanies official
NBER publications.
© 2012 by Richard Hornbeck and Suresh Naidu. All rights reserved. Short sections of text, not to exceed
two paragraphs, may be quoted without explicit permission provided that full credit, including © notice,
is given to the source.
When the Levee Breaks: Black Migration and Economic Development in the American South
Richard Hornbeck and Suresh Naidu
NBER Working Paper No. 18296
August 2012
JEL No. N32,N52,O10
ABSTRACT
In the American South, post-bellum economic stagnation has been partially attributed to white landowners'
access to low-wage black labor; indeed, Southern economic convergence from 1940 to 1970 was associated
with substantial black out-migration. This paper examines the impact of the Great Mississippi Flood
of 1927 on agricultural development. Flooded counties experienced an immediate and persistent out-migration
of black population. Over time, landowners in flooded counties dramatically mechanized and modernized
agricultural production relative to landowners in nearby similar non-flooded counties. Landowners
resisted black out-migration, however, benefiting from the status quo system of labor-intensive agricultural
production.
Richard Hornbeck
Department of Economics
Harvard University
232 Littauer Center
Cambridge, MA 02138
and NBER
hornbeck@fas.harvard.edu
Suresh Naidu
Columbia University
School of International and Public Affairs
MC 3328
420 West 118th Street
New York, NY 10027
and NBER
sn2430@columbia.edu
Under-developed societies often have a large population of low-wage agricultural workers.
Economic growth requires a reallocation of labor, yet various factors may keep workers
in rural agriculture (Lewis, 1954; Kuznets, 1955; Brenner, 1986; Banerjee and Newman,
1998). Low-wage agricultural labor may discourage labor-saving technological innovation
(Habakkuk, 1962; Allen, 2009; Acemoglu, 2010) or the adoption of new capital-intensive
technologies (Atkinson and Stiglitz, 1969; Basu and Weil, 1998).
The Southern United States experienced a remarkable economic transition from 1940 to
1970 (Wright, 1986). The US South shifted from a largely agrarian low-wage economy to a
more industrial economy paying comparable wages to the North; within the agricultural sector, production became more capital-intensive and farm sizes increased. The mechanization
and modernization of Southern agriculture coincided with large-scale black out-migration,
though a direct causal relationship is difficult to observe.
At the beginning of the 20th century, Southern white planters dominated areas with
concentrated black populations. The Mississippi Delta exemplified this system of racial
inequality and discrimination that fostered paternalistic black labor relations and narrowed
black economic opportunities. The Great Mississippi Flood of 1927 displaced workers and
disrupted the traditional racial labor market equilibrium, leading to an exodus of black
laborers and sharecroppers from flooded areas.
This paper examines the impact of the 1927 Mississippi flood on agricultural development,
emphasizing the relationship between black out-migration and Southern economic development. Empirical estimates support historical accounts of a black exodus from flooded areas.
Agriculture then became substantially mechanized and modernized in flooded counties relative to nearby similar non-flooded counties. Estimated changes in agricultural land values
are consistent with white landowners’ coercive efforts to resist black out-migration after the
flood and maintain the status quo system of labor-intensive agricultural production.
Using county-level data from the Censuses of Agriculture and Population, from 1900 to
1970, the main empirical specifications compare changes between flooded counties and nonflooded counties within the same state and with similar pre-1927 outcome values. The
analysis of black population declines is supplemented with individual-level Census data,
matched between 1920 and 1930.
The empirical estimates are robust to controlling for other differences between flooded and
non-flooded counties, including differential changes associated with: distance to the Mississippi river; geographic suitability for cotton and corn; terrain ruggedness; or longitude and
latitude. The estimates are also robust to controlling for differential intensity of plantations
or differential impacts of New Deal program spending.
In a similar analysis of counties near other major Southern rivers, compared to counties
1
further from rivers, there is little estimated relative change in black population or agricultural
development in the absence of a flood. Counties near other major Southern rivers exhibit
many of the same outcome patterns prior to 1927, yet do not experience the subsequent
large relative changes estimated in flooded counties.
Our main interpretation of the empirical results is that flood-induced black out-migration
encouraged the adoption of capital-intensive technologies and larger-scale farm operation,
consistent with contemporary and historical qualitative accounts. The empirical estimates
appear less consistent with alternative interpretations, such as direct impacts of the flood on
capital investment or land productivity. Further, general equilibrium impacts on non-flooded
counties appear to be small.
In the wake of the 1927 flood, black out-migration and agricultural development in flooded
areas provides a microcosm of subsequent out-migration and development across the American South. In under-developed societies with a substantial population of low-wage agricultural workers, rural out-migration may encourage agricultural mechanization and modernization. Whether caused by push factors (e.g., rural natural disasters) or pull factors
(e.g., urban labor demand), decreased agricultural labor surpluses may promote structural
economic development.
I
Historical Background
I.A
Southern Under-development and the Mississippi Delta
Even prior to the revolutionary war, the Southern economy was distinctive. Slavery and a
geographic suitability for plantation agriculture contributed to a system of labor-intensive
agricultural production. As slavery expanded into new states during the 19th century, political conflict between Northern free states and Southern slave states culminated in the Civil
War. Four million slaves were emancipated and enfranchised; by 1900, however, most Southern states had effectively disenfranchised black populations via poll taxes and literacy tests
(Naidu, 2012).
Southern white planters attempted to use their political influence to restrict black labor
mobility and exert control over black agricultural workers.1 Anti-enticement laws made it
illegal for one planter to hire another planter’s workers, while anti-vagrancy laws made it
illegal to be unemployed and without housing (Naidu, 2010). There has been substantial
debate over the effectiveness of these measures and the overall degree of black labor mobility
(see, e.g., Myrdal, 1944; Higgs, 1973; Mandle, 1978; Wright, 1986; Fishback, 1989; Margo,
1990, 1991; Ransom and Sutch, 2001; Alston and Kauffman, 2001); less controversial, how1
We use agricultural “workers” to refer to both wage laborers and tenant farmers, who received “wages”
in the form of production shares, housing, and advances of inputs and/or money.
2
ever, is that Southern white planters valued black labor immensely and used both carrots
and sticks in an attempt to retain labor.
Southern black labor relations were also distinguished by the threat of racial violence
(Rosengarten, 1975; Tolnay and Beck, 1995). Southern white planters often pursued a strategy of paternalism to retain black workers, offering protection from white violence and implicit insurance. “Protection was important .... particularly for black workers, because they
lacked civil rights and society condoned violence”(Alston and Ferrie, 1999, p. 20). During a
period of labor scarcity, a team of anthropologists observed: “One of the bases of competition between landlords for tenants was the landlord’s reputation among tenants with regard
to his use of physical violence. At the same time the field evidence reveals that the use of
threats of violence by white planters is one of the basic controls upon labor” (Davis, Gardner
and Gardner, 2009, p. 392).
The Southern economy remained persistently under-developed between the Civil War and
World War II, and the relative abundance of low-wage black agricultural labor is one potential
explanation. While the North developed large manufacturing sectors, the South remained
primarily agricultural. Northern wheat threshing became increasingly mechanized in the
19th century (David, 1975), while the mechanization of Southern cotton-picking was delayed
until the mid-20th century.2
Early cotton mechanization was mainly in planting and cultivation, where replacing mules
and horses with tractors was associated with a 30% reduction in labor inputs (Hurst, 1933).
Tractors and other labor-saving innovations have been influential in American agricultural
development (Olmstead and Rhode, 2001; Gardner, 2002; Steckel and White, 2012), yet
adoption lagged in the South. “Technology for mechanizing the preharvest operations was
available well before the 1930s, yet it was hardly used at all in the South, and least of all in
the plantation belt” (Wright, 1986, p. 133). Early tractors could replace mule-drawn carts in
transporting cotton to gins (Ellenberg, 2007), yet continued high demand for harvest labor
encouraged annual labor contracts and may have discouraged the partial mechanization of
pre-harvest operations (Whatley, 1987). “Not only cheap labor, but also the form of that
cheap labor, reduced the profitability of mechanization” (Whatley, 1985, p. 1208).
Southern economic convergence and agricultural mechanization coincided with large-scale
black out-migration. Wright (1986) describes a 1940 to 1970 economic transition from the
“Old South” to the “New South,” attributing much of this change to a breakdown of regional
labor markets and increased mobility of Southern blacks out of the region. Contemporaries
recognized a feedback relationship between labor scarcity encouraging agricultural mecha2
These differences may not just reflect crop choice, as California mechanized cotton before the US South
(Whatley, 1985).
3
nization and technological improvements displacing workers (Raper, 1946).3 Farm sizes increased as agriculture became more capital-intensive and as mules and horses were replaced
with tractors and harvesters (Kirby, 1987).4
The United States’ Southern economy experienced remarkable economic growth in the
mid-20th century. Much regional convergence in the United States was driven by labor
movement out of Southern agriculture and relative increases in Southern agricultural wages
(Caselli and Coleman, 2001). Aside from the role of black out-migration, however, important
events were the New Deal, World War II, and Civil Rights regulation (Wright, 1986; Heckman
and Payner, 1989; Donohue and Heckman, 1991; Besley, Persson and Sturm, 2010).5
This paper focuses on the lower Mississippi region, which embodied historical Southern
under-development. The Mississippi-Yazoo Delta has been dubbed the “most southern place
on earth” (Cobb, 1994), and became infamous for racial inequality and abuse.6 However,
powerful white planters recognized their economic dependence on local black labor. Some
planters experimented with recruiting Chinese and Italian workers, but were unable to find
adequate and willing substitutes. Planters, such as Leroy Percy, resisted the Klu Klux Klan
to protect their black workers. Retaining a local labor force became increasingly difficult,
however, during World War I and the first Great Migration. The lower Mississippi region
would soon experience an exodus of black labor that foreshadowed the second Great Migration across the US South.
I.B
The Great Mississippi Flood of 1927
“A great deal of labor from the flooded section after being returned to the plantations is going North. It is thus a serious menace and it is going to offer a
tremendous problem to all of us” — Alex Scott, Delta planter.
The Mississippi river basin stretches into the central United States to channel water down
through the winding Mississippi river. The river itself is somewhat undefined, historically
changing course and spilling into natural floodplains. Over the late 19th and early 20th
centuries, levees were constructed to contain the river and its natural spillways were closed.
In 1926, the new chief of the Army Corps of Engineers “for the first time officially stated in
his annual report that the levees were finally in condition ‘to prevent the destructive effect
3
Some economic historians have emphasized the role of the mechanical cotton picker in displacing workers (Day, 1967; Grove and Heinicke, 2003), while others have emphasized the impact of labor scarcity on
mechanization of the cotton harvest (Peterson and Kislev, 1986; Holley, 2000).
4
While land ownership was often concentrated, “farm size” refers to the parcel size of farm operators.
5
Additional important factors include malaria eradication (Bleakley, 2007) and the introduction of air
conditioning (Arsenault, 1984).
6
In 1921, William Pickens, Arkansan NAACP secretary, dubbed the Mississippi River Valley the “American Congo.” In 1919 alone, at least 18 black citizens were lynched in the Delta (Woodruff, 2003).
4
of floods”’ (Barry, 1998, p. 175).
In 1927, the levee system failed catastrophically along the lower Mississippi river. Heavy
rains throughout the Mississippi river basin accumulated in rising river levels and enormous
pressure created 145 levee breaks that flooded 26,000 square miles. In the three most-affected
states (Mississippi, Louisiana, Arkansas), flooding hit 36% of agricultural land and 29% of
the population (Red Cross, 1928). The flood is estimated to have caused $400 million in
property damage and drowned 246 people.7
The Red Cross coordinated flood relief efforts, which focused on emergency short-term
needs (Red Cross, 1928). Of the $17 million spent, 30% was for food and 14% was for
livestock feed. The Red Cross spent 16% on seed for farmers to replant flooded cropland:
two-thirds of this land could be replanted in 1927, though the late planting season required
some land to be shifted from cotton to corn, and the remaining lands were replanted in 1928.
Building construction, repairs, and household furnishings totaled 15% of expenditures, and
the remaining 25% was mainly for rescue and setup of refugee camps.8
The Red Cross established refugee camps that held 45% of the black population from
flooded areas in Mississippi, Louisiana, and Arkansas (Red Cross, 1928).9 Refugee camp
administration was placed under the control of local counties and, in effect, powerful local
white planters.
Many refugee camps became centers of repression and racial abuse. Black work gangs
were conscripted and forced to work on levees or planters’ farms; those caught attempting
to leave were beaten and returned.10
Flood relief mainly refused direct payments to individuals, instead providing in-kind transfers through Red Cross camps.11 Much of this aid was captured by white planters or withheld unless blacks worked on prescribed tasks. Planters justified withholding rations on the
grounds that rations would “spoil” black workers and weaken the control planters had in
“the old system” (Spencer, 1994, p. 176).
Amidst stories of racial abuse, white planters in flooded areas retained little credibility in
7
There was little flood insurance at this time (White, 1945), and “[f]looding in the MS basin in 1927 and
1928 led the few companies that were selling cover to abandon the business. It was not until the 1950s that
flood insurance again began to be discussed seriously” (Parker, 2000, p. 413).
8
An additional $6 million in services and supplies were donated by the railroads, US military, and other
Federal agencies, mainly for rescue and setup of refugee camps.
9
Refugee camps held 26% of the white population from flooded areas in these three states. The Red
Cross also gave relief outside of camps to 33% of the white population and 36% of the black population from
flooded areas.
10
In May 1927, 21 black workers were caught and whipped by the National Guard for trying to escape
a relief camp (Spencer, 1994, p. 177). In another case, a black insurance officer who refused to work was
openly shot and killed by the mayor of Lake Providence, LA (Barry, 1998, p. 330).
11
The Bill prefigured New Deal legislation by providing a federal transfer to landowners without requiring
local contributions.
5
offering paternalistic protection to their black workers. One infamous Red Cross camp in
the Delta was controlled by Will Percy, son of LeRoy Percy, who forced blacks to work in
the camp for free and wear laborer tags to receive food.12 “Following a killing of a black man
by a white policeman on the levees, Will Percy gave a condescending lecture to the black
community at Mount Horeb church ‘Because of your sinful, shameful laziness, because you
refused to work on your own behalf unless you were paid, one of your race has been killed.’
After this, the bond between the Percys and the blacks was broken” (Barry, 1998, p. 333).
A circulated black newspaper, The Chicago Defender, provided detailed accounts of racial
abuse in Red Cross camps and listed job openings for blacks in Northern cities.13 Migration
costs also declined as a result of the flood, due to temporary displacement and lower labor
demand for cotton harvesting in 1927.
Faced with the potential exodus of black workers, white planters made every effort to
retain their black labor force. Following directives from the Mississippi governor and the
National Guard commander, the Red Cross issued a memo on the “return of refugees,”
stating: “Plantation owners desiring their labor to be returned from Refugee Camps will
make application to the nearest Red Cross representative,” whereupon they “will issue passes
to refugees” (Barry, 1998, pp. 313-314). The Delta & Pine Land Company, one of the nation’s
largest cotton plantations, established its own refugee camp and had its workers transferred
by special train.
Despite such efforts, or perhaps encouraged by such efforts, many black families left flooded
areas in search of better political and economic opportunities. Contemporary accounts describe black families, once displaced from their homes, continuing on to Chicago and other
Northern cities.14 “The Afro-American reported that the relief camp experience had “inspired many backwoods farm[h]ands to to make their first break for better things”’ (Spencer,
1994, p. 177). Social networks shifted toward favoring migration; in Greenville MS, black
leaders left for Chicago and crowds of blacks gathered at the local railway station every
Saturday night to see who was leaving and say goodbye (Barry, 1998).15
12
Barry (1998, p. 315) recorded a black man saying: “The colored people caught tough times around
Greenville.... Whites were kicking coloreds and beating them and knocking them around like dogs. Hungry
people, they wouldn’t feed them sometimes.” A white woman remembered: “The [National] Guard would
come along and say ‘There’s a boat coming up. Go unload.’ If they didn’t hurry up, they’d kick them. They
didn’t mind taking their guns, pistols out, and knocking them over the head.”
13
Commerce Secretary Herbert Hoover gained national prominence through his management of flood relief
operations and secured the presidential nomination. However, racial abuses during the flood eventually cost
him the support of national black leader Robert Moton, who had been in charge of investigating racial abuses
in relief camps, and contributed to the departure of blacks from the Republican party.
14
This episode was influential in the development of Delta blues and Chicago blues (see, e.g., “When the
Levee Breaks” and “High Water Everywhere”).
15
Reverend E.M. Weddington, who pastored Mount Horeb church, left shortly after the flood receded,
but not before allegedly writing an anonymous letter saying “All of this mean and brutish treatment of the
6
Landowners’ accounts emphasize the damages from losing their labor force, rather than
direct losses from the flood. LeRoy Percy reported: “The most serious thing that confronts
the planter in the overflowed territory is the loss of labor, which is great and is continuing”
(Barry, 1998, p. 416). The director of the Delta Land and Pine Company reported to shareholders: “Labor was completely demoralized and the plantation was left almost completely
without labor.”
White planters in flooded counties were forced to adapt to the decreased availability of
black workers. In November 1927, the Engineering News Record noted: “In certain sections
of the lower Delta above the Arkansas and Yazoo where a crop could not be made this year
two-thirds to four-fifths of the families have moved away. In these districts farm-machinery
salesmen have been busy, and farm experts are watching the result with some apprehension.”
In 1931, a Mississippi Agricultural Extension Service bulletin discusses the “serious problem”
of black out-migration and explores “the possible solution in mechanical farming,” comparing
five tenant-operated plantations and five tractor-operated plantations in the Delta (Vaiden,
Smith and Ayres, 1931). Contemporaneous accounts describe a reorganization of agricultural
production and increased mechanization in the Delta, even prior to the introduction of the
cotton harvester: “Many planters have turned to the use of wage labor and large-scale
machinery in an effort to improve production efficiency and decrease costs” (Langston and
Thibodeaux, 1939, p. 3).
The Mississippi Delta has often been examined as a microcosm of historical Southern
underdevelopment; after the 1927 flood, the Delta also provides a microcosm of Southern
economic development following black out-migration. In contrast to the subsequent black
out-migration across the South, the particular flooded areas experienced a sharp exogenous
decrease in black population due to temporary displacement and a decline in the opportunity
cost of migration, a breakdown of trust between planters and black workers, and a shift in
black social networks toward favoring migration.
II
A Model of Flooding, Migration, and Agricultural Development
II.A
Model Setup
Assume that a representative Southern planter in county  and year  produces agricultural

goods for a world market with fixed prices:   (  
   ). Each county has a fixed
supply of land with productivity  . Capital  is defined broadly to include equipment
and machinery, mules and horses, fertilizer, and land improvements. Capital is sufficiently
mobile or depreciable that the marginal return to capital  is equalized across counties. Labor
colored people is nothing but downright slavery” (Barry, 1998, p. 416).
7
 16
is supplied inelastically by resident black workers 
 and resident white workers  .
Capital and labor are assumed to be substitutes, reflecting a choice between “Old South”
labor-intensive production and “New South” capital-intensive production.17 Capital is an
important input in older production methods, but newer production methods are embodied
in capital goods. Black workers and white workers are also substitutes, and we consider
allowing for higher capital-labor substitutability for black workers (e.g., due to differences in
average education).
White workers are perfectly mobile and earn a fixed outside “Northern” wage normalized

. Planters
to  . Black workers can earn an outside wage  or a home county wage 
have established a reputation for protecting their own workers from racial violence, which
is worth  to black workers in each period. Black workers also pay a one-time moving cost
, equivalent to paying  in each future period, reflecting racially-biased labor market

=  −  − . Each
institutions. Home county wages are set in equilibrium such that 
county is in an initial steady-state with 0 black workers.

In the first period, the Southern planter chooses inputs to maximize:   (1  
1  1 ) −
 

 18
We focus on the case in which this
1 − ( −  − )
1 −  1 , subject to 1 ≤ 0 .


constraint binds and 1 = 0 , consistent with efforts by Southern planters to limit black
out-migration. Capital investment and the number of white workers are determined by:
(1)
(2)

  (1  
0  1 ) = 


  (1  
0  1 ) = 
In particular, equilibrium choices of capital and white workers depend on the initial number
of black workers: more black workers leads to a lower capital stock and fewer white workers.
Paternalism and moving costs both have the effect of lowering planters’ labor costs in counties
with more black workers.
II.B
Comparative Statics after the Flood
Consider the impact of a flood in some counties between periods 1 and 2. Workers are
housed in refugee camps controlled by the planter, and racial abuses in these camps lower
16
“Workers” include wage laborers, share croppers, and share tenants who receive “wages” in the form of
cash, production shares, housing, and/or inputs.
17
In particular, we assume that the above production function represents an upper envelope over “Old
South” and “New South” technological choices. Increased use of “New South” methods is assumed to be
“strongly labor-saving” (Acemoglu, 2010); that is, in the case where machines replace labor, the adoption
of capital-intensive methods reduces the marginal product of labor. Note, however, that output per worker
will still increase following a decline in labor availability and the adoption of labor-saving technology.
18
The planter could hire more black workers at wage  , but this would contradict the assumption of an
initial steady-state with 0 black workers.
8
the planter’s ability to provide credible protection from white violence. The flood also
temporarily reduces the moving cost for black workers, either by imposing some share of
that cost or by reducing the opportunity cost of migration. Black workers in refugee camps
may also receive additional information about Northern job opportunities or, as leaders of
the black community migrate, social networks may shift toward encouraging migration. The
value of protection falls to some fraction  and the cost of moving falls to some fraction ,
though the planter may use a combination of incentives and threats to induce workers to
return at cost (1 − ) + (1 − ).

After the flood, the Southern planter chooses inputs to maximize:   (2  
2  2 ) −
 


2 − ( −  − )
2 −  2 , subject to 2 ≤ 0 . The flood effectively increases
the cost of employing black workers. Assume that the flood’s impacts are sufficiently large,
i.e.,  and  are sufficiently small, that the constraint no longer binds and the population of

black workers declines in equilibrium (
2  0 ).
In flooded counties, there will also be increases in the capital stock, the population of white
workers, and output per worker. The loss of low-wage black workers increases planters’ labor
costs, which encourages the adoption of labor-saving capital-intensive production methods.
This technological transition will be especially pronounced if there is a higher substitutability
between capital and black workers; for example, if there is capital-skill complementarity and
white workers are higher-skilled on average.
This model does not include dynamic adjustment costs. In practice, it may take a number
of periods to make technological adjustments and to accumulate the desired capital stock.
However, the out-migration of black workers is predicted to be immediate and persistent.
Agricultural land values reflect the present discounted value of rents and, in this baseline
model, decline immediately due to the loss of exploitable low-cost black labor. For this
reason, land-owning Southern planters are predicted to resist black out-migration. Land
values may increase over time if capital investments become fixed to the land, but this is a
matter of accounting and does not reflect gains for landowners.
If there were sufficiently large externalities in capital investment, however, the flood may
cause a “big push” toward mechanization that increases land values. Allowing for multiple
planters in each county, as a single planter internalizes all within-county spillovers, the private
return to capital investment may be increasing in county-level total capital investment due to
knowledge spillovers or coordinated investments in new capital equipment and infrastructure
(see, e.g., Romer, 1986; Murphy, Shleifer and Vishny, 1989; Foster and Rosenzweig, 1995).
Agricultural mechanization may also increase over time due to learning-by-doing, but land
values would only increase immediately after the flood if there were substantial externalities
associated with anticipated agricultural mechanization.
9
III
Data Construction and Baseline Differences in Flooded Counties
III.A
Data Construction and Aggregate Trends
Historical county-level data are drawn from the Census of Agriculture and the Census of
Population (Haines, 2005).19 The main variables of interest include: black population, value
of agricultural equipment and machinery, number of mules and horses, number of tractors,
average farm size, and value of agricultural land and buildings.20 The value of agricultural
equipment and machinery includes all tools, wagons, cotton gins, threshing machines, and all
other machinery used in carrying out farm business (engines, motors, tractors, automobiles,
and motor trucks); note that this measure excludes the value of mules and horses, levees, or
any land improvements (Census Bureau, 1927).
For the 1920’s, a direct measure of migration is drawn from matched individual-level
census data in 1920 and 1930 (Boustan, Kahn and Rhode, 2012).21 The match rate of
24% is comparable with the existing literature, though false matches will tend to overstate
migration rates. Later analysis examines the fraction of matched individuals in 1930 that
have left their 1920 county, state, or the South (and differences by race).
The empirical analysis focuses on a balanced panel of 163 counties, from 1900 to 1970, for
which data are available in every period of analysis. To account for county border changes,
data are adjusted in later periods to maintain 1900 county definitions (Hornbeck, 2010).
Figure 1 maps the extent of flooding in 1927, overlaid with county borders in 1900. The
shaded area represents the flooded region, as compiled by the US Coast and Geodetic Survey.
To focus the analysis on initially more-comparable flooded and non-flooded counties, the
main sample is restricted to counties with a black population share greater than 0.10 in 1920
and a fraction of cropland in cotton greater than 0.15 in 1920.22 Additional specifications
examine counties elsewhere in the South, particularly those near other major rivers.
Figure 2 reports aggregate changes in the sample region from 1900 to 1970. Black population decreased substantially from 1940 to 1970, during the second Great Migration; and decreased somewhat in the 1910’s, during the first Great Migration (panel A). Total population
increased through 1940, before declining into the 1960’s (panel B). The value of agricultural
capital increased through 1920, remained mainly constant from 1920 to 1940 during a period of relatively few technological improvements, and then increased substantially by 1970
after the second Great Migration, the Civil Rights movement, and introduction of the cotton harvester (panel C). By contrast, the number of mules and horses were mainly constant
19
We thank Michael Haines and collaborators for providing additional data from ongoing collection.
Note that “farm size” refers to the size of farm operator parcels, rather than units of land ownership.
21
We are grateful to Leah Boustan, Matt Kahn, and Paul Rhode for sharing their matched census data.
22
The sample is further restricted to contiguous counties in Arkansas, Louisiana, Mississippi, and Tennessee.
20
10
through 1940, and then declined substantially through 1960 (panel D). Average farm sizes
declined through 1930, before increasing substantially through 1970 (panel E).23 The value
of agricultural land per farm acre increased during World War I, declined somewhat through
the Depression, and then increased substantially through 1970 as agricultural productivity
increased (panel F). This figure provides some background on regional trends, whereas the
main empirical analysis estimates within-state relative changes for flooded counties.
III.B
Baseline Differences in Flooded Counties
In an initial step, the empirical analysis explores pre-differences between flooded and nonflooded counties. In 1925 or 1920, depending on data availability, county outcome  is
regressed on the fraction of county land flooded in 1927 and state fixed effects:
(3)
 =    +  + 
For each outcome variable, the estimated  reflects within-state differences in pre-flood
characteristics for flooded counties and non-flooded counties.
To explore differences in pre-trends between flooded and non-flooded counties, equation
(3) is modified to regress the change in outcome  from 1910 to 1920 (or from 1920 to 1925)
on the fraction of county land flooded in 1927 and state fixed effects:
(4)
 − (−1) =    +  + 
For each outcome variable, the estimated  reflects within-state differences in pre-flood trends
in characteristics for flooded counties and non-flooded counties.
Table 1, column 1, reports average county characteristics prior to the 1927 flood. Column
2 reports within-state differences in pre-flood characteristics for flooded counties, and column
3 reports these differences conditional on six county-level controls (distance from the Mississippi river, geographic suitability for cotton and corn, terrain ruggedness, and longitude and
latitude). Column 4 reports within-state differences in pre-flood trends for flooded counties,
and column 5 reports these differences in trends conditional on the above six county-level
controls.
Prior to the 1927 flood, flooded counties and non-flooded counties are estimated to have
had similar changes in most outcomes.24 Flooded counties had an initially higher black population and a greater intensity of small-scale agricultural production, though these differences
23
Note that increases in “farm size” refer to increases in the size of farm operator parcels, rather than the
concentration of land ownership.
24
The main results tables report pre-flood changes for each outcome variable in all available pre-periods,
relative to 1920.
11
are partly mitigated by the six county-level controls. To the extent that flooded counties
were different in pre-trends or levels, the main empirical specifications report robustness to
controlling for pre-flood differences.
IV
Empirical Framework
The main empirical specifications estimate year-specific differences between flooded counties
and non-flooded counties, relative to a base year of 1925 or 1920. Outcome  in county 
and year  is regressed on the fraction of county land flooded in 1927, state-by-year fixed
effects, and county fixed effects:
(5)
 =      +  +  + 
Note that  is allowed to vary by year, so each estimated  is interpreted as the average
difference between flooded counties and non-flooded counties in that year relative to the
omitted base year of 1925 or 1920. The main identification assumption is that flooded
counties would have changed similarly to non-flooded counties in the same state, if not for
the flood.
In practice, further specifications control for county characteristics ( ) that may predict
differential changes between flooded and non-flooded counties:
(6)
 =      +  +  +   + 
Most specifications control for pre-flood values of the outcome variable, flexibly allowing for
convergence over time in the outcome variable or otherwise differential changes associated
with initially different values (i.e., columns 2 and 3 of Table 1).25 Further specifications
control for distance to the Mississippi river; geographic suitability for cotton and corn; terrain
ruggedness; and longitude and latitude.
For the statistical inference in all specifications, standard errors are clustered at the county
level to adjust for heteroskedasticity and within-county correlation over time. When allowing for spatial correlation among sample counties, the estimated standard errors generally
increase by less than 15%.26 The regressions are weighted by county size, so the estimates
reflect changes for an average acre of flooded land.
25
Note that this specification is not a lagged dependent variable model; instead, the specification controls
only for pre-treatment values of the dependent variable.
26
Spatial correlation among counties is assumed to be declining linearly up to a distance cutoff and zero
after that cutoff (Conley, 1999). For distance cutoffs of 50 miles, 100 miles, or 200 miles, the estimated
Conley standard errors are generally less than 15% higher than the standard errors when clustering at the
county level, depending on the outcome variable and year.
12
V
Main Results
V.A
Population
Table 2 reports estimated changes in black population for flooded counties, relative to changes
for non-flooded counties. From estimating equation (5), column 1 reports that flooded counties experienced a 14% (0.151 log point) decline from 1920 to 1930 in their black population
share. Following the 1927 flood, this short-run decline in black population share persisted
through 1970.
Consistent with the main identification assumption, the black population share changed
similarly in flooded counties and non-flooded counties prior to the 1920’s. Further, from
estimating equation (6), column 2 reports that the estimated decline in black population
share is robust to controlling for changes correlated with counties’ black population share in
1920, 1910, and 1900. The demographic shift was mainly caused by a decline in the black
population (columns 3 and 4), with little change in total population (columns 5 and 6).27
Changes in county-level population reflect net migration, but a more direct measure of outmigration uses matched individual-level census data from 1920 and 1930. Average migration
rates may be overstated due to false matches, though any bias should not be differential
across flooded counties and non-flooded counties.28 The county-level out-migration rate is
calculated as the number of matched people leaving the county, divided by the total number
of matched people originally in the county. The estimated regression is the same as equation
(5), except the regression is weighted by the number of matched people in each county.29
Table 3, panel A and column 1, reports that the fraction of matched people leaving their
county between 1920 and 1930 is 11.8 percentage points higher in flooded counties than in
non-flooded counties. Flooded counties also have a higher fraction of matched people leaving
their state (column 2), though a similar fraction leaving the South entirely (column 3).30
Migration estimates are more striking for the subsample of individuals whose race is observed. Panel B reports estimated differences in black out-migration from flooded counties,
and panel C reports estimated differences in white out-migration from flooded counties.
Blacks in flooded counties are 13.9pp more likely to leave their county (column 1), 17.8pp
more likely to leave their state (column 2), and 6.8pp more likely to leave the South en27
Predicted changes in total population depend on the functional form of the production function, but
the offsetting increase in white population may reflect an inelastic demand for labor. Estimated increases in
farmland (below) imply a decline in population per farm acre.
28
All successful matches are required to be unique by name and place of birth (state or country) within a
5-year age band.
29
In weighting by the number of matched people, the regressions estimate the change in probability of
migration for the average person.
30
Southern states are defined as Arkansas, Louisiana, Tennessee, Mississippi, Alabama, North Carolina,
South Carolina, Georgia, and Florida.
13
tirely (column 3); by contrast, there are no statistically significant differences in whites’
out-migration rates.
Overall, the estimates are consistent with historical accounts of an immediate and persistent decline in black population in flooded counties. The empirical results do not identify
whether this decline in population reflects the flood’s temporary displacement effect and a
decline in the opportunity cost of migration, a breakdown of trust between planters and
black workers, or a shift in black social networks toward favoring migration. Regardless of
the mechanism, however, the subsequent empirical analysis explores the impact of decreased
black labor availability on agricultural development.
V.B
Agricultural Mechanization and Modernization
Table 4, columns 1 and 2, report that the value of agricultural capital equipment and machinery fully recovered in flooded counties by 1930 from losses sustained during the 1927 flood.
By 1940, the value of agricultural capital had increased substantially in flooded counties,
relative to non-flooded counties. Relative increases in agricultural capital continued through
1970; that is, the shift toward mechanization does not simply reflect earlier mechanization
in flooded counties and convergence over time. Figure 3 graphs the estimated coefficients
from column 1 of Tables 2 and 4, highlighting the relationship between a declining black
population share and increased agricultural mechanization.
Mules and horses were an important early source of agricultural power; used by agricultural
workers, but overall a substitute for manpower.31 Columns 3 and 4 report that flooded
counties experienced an initial increase in mules and horses, despite animal deaths as a
direct consequence of the flood. By the 1950’s and 1960’s, however, use of this “Old South”
power source declined.
Tractors were still rare in the sample region during the 1920’s and 1930’s, and estimated
changes are more sensitive to controlling for counties’ number of tractors in 1925. Column
5 reports that flooded counties adopted tractors faster in the late 1930’s and 1940’s, while
column 6 indicates a more permanent increase in tractor usage.32
Increased farm sizes were strongly associated with a transition from an “Old South” system
to a “New South” system of agricultural production. Columns 7 and 8 report that flooded
counties experienced a gradual and substantial increase in average farm size, relative to
non-flooded counties. Farm sizes increased particularly during the 1950’s and 1960’s as
mechanized cotton harvesters became increasingly available.
31
Mules and horses are a form of “capital,” but their value is not included in the value of agricultural
capital equipment and machinery.
32
While tractor quality is unobserved, higher agricultural capital in later periods and a similar number of
tractors may indicate higher tractor quality in flooded counties.
14
It is difficult to measure the increase in labor productivity associated with reported changes
in production inputs and methods. As a proxy, however, data are available for the value of
crops per capita. From estimating equations (5) and (6), the log value of crops per capita
changed similarly in flooded and non-flooded counties from 1910 through 1930. This proxy
for average labor productivity increased substantially in flooded counties through the 1930’s,
1940’s, and 1950’s.33
Overall, the estimated increases in farm capital appear to embody labor-saving technological change in the agricultural sector. Much of the early labor-saving adjustment relied on
mules and horses, i.e., older vintage technologies, though subsequent mechanization appears
to reflect increased use of new labor-saving technologies like the cotton harvester. Gradually
increased farm operation sizes are associated with a bundle of production methods associated
with the “New South” rather than the “Old South.”
The flood itself was likely too small to encourage labor-saving technological innovation
(e.g., Habakkuk, 1962; Allen, 2009; Acemoglu, 2010), but decreased labor availability appears to have made flooded counties more suitable for the adoption of new capital-intensive
technologies (e.g., Atkinson and Stiglitz, 1969; Basu and Weil, 1998). Increases over time in
agricultural mechanization and modernization may reflect learning-by-doing, increased availability of wage workers during the Depression and New Deal, and/or subsequent increases
in the availability of cotton harvesters and other mechanical innovations.
V.C
Farmland Acreage and Value
Table 5, columns 1 and 2, report that flooded counties experienced a substantial relative
increase in farmland from 1930 through 1970. As farms became larger and more capitalintensive, agricultural production in flooded counties also became more land-intensive. Clearing and plowing additional farmland appears to be complementary with increased mechanization.34
Substantial increases in total farmland, along with increased investment, complicate an
analysis of the value of agricultural land and buildings. In principle, changes in agricultural
land values reflect the loss (or gain) to landowners from decreased labor availability and
subsequent agricultural adaptation. New farmland may be of generally lower quality than
initial farmland, however, causing a downward bias in the value of farmland per farm acre.
33
From equation (5), the estimated coefficients (and standard errors) are 0.395 (0.078) in 1940, 1.235
(0.167) in 1950, and 2.158 (0.246) in 1960. From equation (6), the estimated coefficients are 0.373 (0.067)
in 1940, 0.933 (0.138) in 1940, and 1.741 (0.223) in 1960. Estimated relative changes are similar for the log
value of crops per person living in rural areas of the county.
34
The increase in agricultural land may represent a decrease in land under the public domain, or an increase
in the fraction of privately-owned land that is in operation (or fallow) and captured by Census enumerators.
Note that the empirical specifications estimate relative changes, so the reported increases may also reflect
less of a decline in farmland in some flooded counties relative to non-flooded counties.
15
By contrast, clearing and plowing new farmland requires substantial sunk investments; as
these investments are capitalized into land values, there will be an upward bias in the value
of farmland per county acre.
Flooded counties experienced a substantial and persistent decline in the value of agricultural land and buildings per farm acre (Table 5, column 3). Land values declined further
over time, which may reflect a compositional decline in average land quality. Controlling
for initial differences, flooded counties experienced only marginally statistically significant
declines in land value (column 4).
Flooded counties experienced little immediate change in the value of agricultural land and
buildings per county acre (column 5). Increased land values in later periods may indicate
that landowners unexpectedly benefited from technological innovation that favored capitalintensive agricultural production; however, rising land values also reflect substantial sunk
investments in clearing and plowing new farmland (column 6). Across all four specifications,
the estimates reject a substantial immediate increase in agricultural land values that might
suggest landowners anticipated benefiting from the forced economic transition.35
Landowners’ coordinated resistance to black out-migration is consistent with landowners
not anticipating economic gains from a “big push” toward increased agricultural mechanization. Indeed, Appendix Figure 1 shows that the Delta Land and Pine Company did not
experience an increase in reported profits (Dong, 1993).36 Overall, the estimates appear
consistent with a single equilibrium in which landowners adapt to labor availability.
Overall, this empirical exercise is not focused on whether the flood itself was beneficial
or detrimental to landowners and workers. Black migrants presumably benefited from the
option to migrate after the flood, though this does not imply that migrants benefited overall
from the flood.37 Instead, the empirical exercise is focused on potential lessons from this
historical episode for understanding the relationship between decreased labor availability and
increased agricultural mechanization and modernization.
35
Data on land values and building values are available separately, by decade, from 1900 to 1940. In 1920,
the value of land averages 77% of the combined value of land and buildings. Focusing on changes in the
value of land only, in 1930 and 1940, the estimates from columns 3 and 5 are more negative and statistically
significant and the estimates from columns 4 and 6 are more negative and statistically insignificant.
36
The Company’s return on investment likely declined, as profits remained similar and capital investment
increased. The Delta Land and Pine company, however, was special in that given its size, it was reluctant to adopt labor-saving machinery due to public disapproval of the resulting unemployment. While it
experimented with tractors in the early 1930s, it did not adopt them until World War 2.
37
There would need to be some externality or coordination failure among migrants to generate welfare
gains from flood-induced out-migration when the black population had previously chosen to stay in the
region. Chay and Munshi (2012) examine Southern black migration networks in the early 20th century,
which are consistent with potential externalities.
16
VI
Threats to Validity
VI.A
General Robustness
An empirical concern is that inherent differences between flooded and non-flooded areas may
have caused some county characteristics to change differently after 1927, even in the absence
of the flood.38 A series of robustness checks explore the importance of inherent differences
between flooded and non-flooded counties. In Tables 6 — 9, the baseline results are robust
to a variety of control variables and alternative specifications.
As a basis for comparison, column 1 presents the baseline results when controlling for initial
outcome differences. Table 6, panel A, reports estimated changes in black population share,
relative to 1920; panel B reports estimated changes in the value of agricultural machinery and
equipment, relative to 1925. Table 7 reports estimated changes for black population (panel
A) and total population (panel B). Table 8 reports estimated changes for mules and horses
(panel A), tractors (panel B), and average farm size (panel C). Table 9 reports estimated
changes for farmland (panel A), the value of farmland per farm acre (panel B), and the value
of farmland per county acre (panel C). Some years’ coefficients are omitted for conciseness.
Column 2 controls for counties’ distance to the Mississippi river, interacted with each year.
Counties closer to the Mississippi are more likely to be flooded in 1927, and nearby counties
have greater historical flooding and better river access to markets. This specification allows
for the impact of river proximity to change over time. Alternatively, the estimates are robust
to restricting the sample to counties within 50km or 100km of the Mississippi river.
Column 3 controls for counties’ geographic suitability for cotton and corn, separately
interacted with each year. Cotton and corn are the two major crops in 1925 in the sample
region. Crop suitability reflects the maximum potential yield of that crop, as calculated
by the FAO using data on climate, soil type, and ideal growing conditions for that crop.39
This specification allows for crop-specific changes in technology and prices, or changes that
otherwise differentially affect areas suitable for different crops.
Column 4 controls for counties’ terrain ruggedness, interacted with each year. Counties’
ruggedness is measured as the standard deviation in altitude across points in the county,
calculated from the USGS National Elevation Dataset.40 Areas with more uniform terrain
may be more suitable for mechanization, or may otherwise change differently over time.41
38
In particular, the assumption that flooded and non-flooded areas would have changed similarly becomes
stronger in later periods.
39
The FAO’s Global Agro-Ecological Zone maps (version 3.0) are used to create county-level average crop
suitability for cotton and corn. Potential yields are calculated using climate averages from 1961 to 1990 and
rain-fed conditions with intermediate inputs.
40
The estimates are similar when ruggedness is measured by the maximum range in altitude across points
in the county.
41
The estimates are also similar when controlling for interactions between terrain ruggedness and geo-
17
Column 5 controls for counties’ longitude and latitude, separately interacted with each
year. This specification controls for spatial patterns in economic changes that may be correlated with flooding.
Column 6 includes all of the controls from columns 2 — 5.
Column 7 instead measures flood intensity using the fraction of population affected by
flooding in each county. Rather than focus on the fraction of each county’s land that is
flooded, these estimates use Red Cross reports on the population affected by flooding in
each county.42
Column 8 controls for counties’ estimated flood propensity score, interacted with each year.
The probability that a county experienced any flooding is modeled as a probit function of
the county’s black population share in 1920 and fraction of cropland allocated to cotton in
1920. The sample is limited to flooded and non-flooded counties with overlapping values of
this propensity score, which removes 6 of the original 163 counties. This specification is an
alternative method to control for initial differences in county outcomes, and to ensure that
flooded and non-flooded counties are drawn from an initially similar sample.
VI.B
Falsification Exercise
Despite the above robustness checks, non-flooded areas may be an inherently poor control for
flooded areas near the Mississippi river. As an alternative check on the results, a falsification
exercise explores whether there are also differential changes between counties close to other
major Southern rivers and counties further from other major Southern rivers. Restricting
the analysis to non-flooded states, this sample includes 171 counties within 50km of a major
river and 72 counties between 50km and 150km of a major river.43
As in the main sample, counties near other major Southern rivers have a higher black
population in 1920 and a greater intensity of small-scale agricultural production in 1925
than counties further from other major Southern rivers. Further, as in the main sample,
these counties had been experiencing similar trends in the county outcomes of interest.44
Table 10 reports that counties near other major Southern rivers changed similarly after
graphic suitability for cotton and corn (and their main effects).
42
Alternatively, the estimates are robust to using Red Cross data on the fraction of agricultural land
flooded or the fraction of total land flooded (Red Cross, 1928). We are grateful to Paul Rhode for sharing
these Red Cross data, which we supplemented.
43
These cutoffs reflect typical distances to the Mississippi for flooded counties and non-flooded counties,
respectively. As in the main sample, the sample is restricted to counties with a black population share greater
than 0.10 in 1920 and a fraction of cropland in cotton greater than 0.15 in 1920. Data are unavailable for
tractors in the entire South, as we only supplemented available data for the main sample region in 1925.
The major rivers shapefile was obtained from ESRI Inc. (“Major Rivers of the United States”).
44
Analogous to the estimates from Table 1, these estimates refer to modified versions of equations (3) and
(4), where the fraction of county flooded is replaced with a dummy variable for whether the county is within
50km of a major river.
18
1927 to counties further from other major Southern rivers.45 Of the few statistically significant estimates, counties close to other major Southern rivers experienced somewhat less
mechanization over time.
Overall, in the absence of a catastrophic flood, counties near other major Southern rivers
do not experience the black out-migration and increased mechanization that appeared in
counties flooded by the Mississippi in 1927. While the Mississippi river is a special river
within the Southern United States, other counties near major rivers showed many of the same
differences in characteristics prior to 1927. These estimates lend support to the identification
assumption that flooded counties would have changed similarly to non-flooded counties in
the absence of the flood.
VI.C
Plantation Counties and New Deal Programs
One particular empirical concern is that flooded counties are more likely to be “plantation
counties,” as recorded by Brannen (1924) for 1910.46 New Deal programs may have displaced
tenants from plantations, increasing the availability of harvest laborers and encouraging preharvest mechanization in the 1930’s (Whatley, 1983; Depew, Fishback and Rhode, 2012).
In non-flooded Southern states, however, plantation counties are estimated to have little
differential change from 1930 to 1940 in the main outcome variables of interest.47 Between
1940 and 1970, plantation counties do experience some relative declines in black population
and increases in agricultural capital, average farm size, and farmland.48
In the main sample of flooded and non-flooded counties, the empirical results are robust to controlling for differential changes in plantation counties.49 Further, the impacts
of the flood on agricultural development are not driven by plantation counties. Allowing
for heterogeneous effects of the flood on plantation counties and non-plantation counties,
the non-plantation counties experience clear declines in black population and increases in
45
In a modified version of equation (6), the fraction of county flooded is replaced with a dummy variable
for whether the county is within 50km of a major river. The specification controls for changes over each time
period that are correlated with state and initial outcome differences.
46
A plantation is defined as a “unified agricultural organization of considerable size under one management,
of practically a continuous tract of land, operated as a single unit with respect to the methods of control of
labor and products, all of which may be worked by wage hands, or all or a part of which may be subdivided
and let to tenants” (Brannen, 1924, p. 9). Brannen used since-lost census data and judgment to select
counties where “plantation farming in these counties is known to be important” (Brannen, 1924, p. 69).
47
In a modified version of equation (6), the fraction of county flooded is replaced with a dummy variable
for whether the county is a “plantation county.”
48
In a modified version of equation (6), as described above, plantation counties have some statistically
insignificant relative declines in black population. Relative to 1925, agricultural mechanization is similar in
plantation counties through 1940 and higher by 0.23 log points by 1970.
49
As in equation (6), the specification controls for a dummy variable for whether the county is a plantation
county (interacted with year).
19
agricultural development.50 The results are also robust to controlling for differential changes
associated with five measures of New Deal spending.51
VI.D
Alternative Interpretations
Consistent with contemporary and historical qualitative accounts, our main interpretation
of the empirical results is that flood-induced black out-migration encouraged agricultural
development. Black labor availability had discouraged the accumulation of agricultural capital and encouraged smaller-scale farm operation. This Southern system of labor-intensive
agricultural production began to break down in flooded areas, previewing the subsequent
breakdown of this system throughout the South over the 20th century.
There are two other main channels through which the 1927 flood may have had lasting
economic impacts. First, the flood may have caused general economic disruption and the
replacement of vintage capital stocks with more technologically-advanced capital.52 Second,
the flood may have changed land productivity.
In the first case, by causing general economic disruption, the flood may encourage landowners to reevaluate and update agricultural production.53 In particular, reconstruction may
replace damaged “vintage” capital goods with newer capital goods, leading to a short-run
increase in capital investment and modernized capital equipment in flooded areas. As capital
stocks depreciate in non-damaged areas, however, natural replacement leads to convergence
in the quantity and age of capital goods.
The empirical results are generally inconsistent with this first alternative interpretation.
The value of agricultural capital equipment and machinery is found to diverge over time in
flooded counties, rather than increase immediately and converge over time. Initial increases
in capital investment were also mainly in older capital goods, such as mules and horses.
Technologically-advanced capital goods, such as cotton harvesters, did not replace older
50
In a modified version of equation (6), the fraction of county flooded is interacted with a dummy variable
for whether the county is a “plantation county” and a dummy variable for whether the county is a “nonplantation county.”
51
As in equation (6), the specification controls for per-capita spending through the AAA, public works,
relief, loan, and guaranteed loan programs (Fishback, Horrace and Kantor, 2005). Note that New Deal
spending is potentially endogenous to the flood, particularly as networks developed by local politicians to
obtain flood relief could be later used to secure New Deal spending.
52
Related alternative explanations are that the flood could have encouraged the coordination and consolidation of land holdings or induced a series of foreclosures that allowed new entrepreneurial farmers to enter.
Land ownership was fairly concentrated and stable in this region, so we do not focus on these related alternative explanations. To the extent that land owners attempted to coordinate investments and production,
this coordination was mainly in maintaining the status-quo labor-intensive system rather than coordinating
over land assembly and increased mechanization.
53
The lower Mississippi region had an unfortunate history of natural disasters in the early 20th century
(Boustan, Kahn and Rhode, 2012); while none were as large as the 1927 flood, this was a volatile region that
appears less likely to have settled into economic complacency.
20
capital goods until well-after the initial reconstruction.
Historically high levels of capital depreciation imply that post-flood capital reconstruction
would have few persistent “vintage capital” effects. While tractors are among the more
durable capital goods, an approximate annual depreciation rate of 12% implies that roughly
85% of investment in 1927 would have depreciated by 1935 (Hurst, 1933). Investment in
agricultural buildings may be more durable; from estimating equation (6), however, the
value of agricultural buildings in flooded counties declined slightly by 1930 and 1940.54
In the second case, by changing land productivity, the flood may directly impact land
values and factor demand. While repeated historical flooding of the Mississippi contributed
to the formation of productive soils, one isolated flood would have limited direct benefits
for soil productivity. The flood also damaged land improvements, but these were generally
rebuilt quickly and substantial new lands were improved and brought under cultivation in
flooded counties.55 It is difficult to know whether the 1927 flood and the subsequent 1928
Flood Control Act increased or decreased landowners’ expected flood risk, though there
should be less differential change in perceived future risk once controlling for distance to the
Mississippi river or limiting the sample to counties near the Mississippi river.56
From estimating equation (6), flooded counties experienced little immediate change in
cotton productivity or corn productivity.57 In subsequent years, cotton and corn acreages
expanded and there was little systematic change in productivity. These estimates are also
consistent with literature on early mechanization being labor-saving but not yield-increasing
(Hayami and Ruttan, 1985).
Finally, for interpreting the main results, the flood may have general equilibrium impacts
on nearby non-flooded counties. The empirical estimates overstate the aggregate impact of
the flood for particular outcomes that are affected oppositely in non-flooded counties. Our
interpretation of the results focuses mainly on the flood’s relative impacts, however, such
as changes in the relative availability of black labor and the relative change in agricultural
mechanization and modernization.
54
Data on land values and building values are available separately, by decade, from 1900 to 1940. The log
value of building values, per farm acre or per county acre, is regressed on the fraction of the county flooded
in 1927, state-by-year fixed effects, county fixed effects, and county outcome values in 1900, 1910, and 1920,
interacted with each year.
55
Red Cross efforts to introduce new varieties of crops and livestock were generally limited (Red Cross,
1928). Reconstruction efforts were focused on emergency needs and temporary relief.
56
The 1928 Act was mandated to protect all of the potentially flooded counties, not just those that
were actually flooded, and thus involved substantial upriver tributaries rather than a sole focus on levees.
Further, reconstruction and modification of the levee system had little direct effect on available agricultural
land, irrigation, or drainage.
57
The log quantity of cotton or corn yielded per harvested acre is regressed on the fraction of the county
flooded in 1927, state-by-year fixed effects, county fixed effects, and county outcome values in 1900, 1910,
1920, and 1925, interacted with each year.
21
The flood may be expected to have little indirect impact on non-flooded counties in subsequent years and decades, even if the flood initially disrupted non-flooded counties. There
may even be small immediate impacts on non-flooded counties’ output prices and return on
capital, given the degree of integration in agricultural markets and the small share of agricultural output directly affected by the flood. As a test of the magnitude of local economic
spillovers, Table 11 reports the estimated change in counties bordering the flooded region,
relative to counties 100km from the flood border.58 Consistent with small local economic
spillovers, particularly in the immediate aftermath of the flood, there was little change in
counties bordering the flooded region compared to further counties.
VII
Conclusion
The Great Mississippi Flood of 1927 was a transformative event in Southern economic history. In a region infamous for oppressive racial institutions, the flood led to an exodus of
black agricultural workers. The resulting decline in black labor availability led to change in
agricultural practices. Agriculture in flooded counties became substantially mechanized and
modernized relative to agriculture in nearby similar non-flooded counties.
The flood imposed immediate direct costs on both white planters and black agricultural
workers, though black workers may have benefited in the long-run from coordinated largescale out-migration. Landowners resisted black out-migration, with physical coercion when
possible, in an effort to maintain labor availability and support a persistent system of laborintensive agricultural production.
The Southern United States experienced a remarkable economic transition from 1940 to
1970, coinciding with large-scale black out-migration. Experiences from the 1927 flood illustrate the role of black out-migration in fostering the mechanization and modernization of
agricultural production; indeed, flooded counties maintained their early lead in mechanization through 1970. In under-developed societies with substantial populations of low-wage
agricultural laborers, rural out-migration may increase agricultural mechanization and modernization. Whether caused by “push factors,” such as rural natural disasters, or caused by
“pull factors,” such as urban labor demand, decreased agricultural labor availability may
promote structural economic development.
58
Each outcome variable is regressed on the (negative) distance from the flooded region in 100km units,
state-by-year fixed effects, county fixed effects, and county outcome values in 1900, 1910, 1920, and 1925
(when available), interacted with each year. An increase in distance from 0km to 100km is equivalent to an
increase from the closest counties to the eightieth centile.
22
References
Acemoglu, D. 2010. “When Does Labor Scarcity Encourage Innovation?” Journal of Political Economy, 118(6): 1037—1078.
Allen, R.C. 2009. The British Industrial Revolution in Global Perspective. Cambridge Press.
Alston, L.J., and J.P. Ferrie. 1999. Southern Paternalism and the American Welfare
State: Economics, Politics, and Institutions in the South, 1865-1965. Cambridge Press.
Alston, L.J., and K.D. Kauffman. 2001. “Competition and the Compensation of Sharecroppers by Race: A View from Plantations in the Early Twentieth Century.” Explorations
in Economic History, 38(1): 181—194.
Arsenault, R. 1984. “The End of the Long Hot Summer: the Air Conditioner and Southern
Culture.” Journal of Southern History, 50(4): 597—628.
Atkinson, A.B., and J.E. Stiglitz. 1969. “A New View of Technological Change.” Economic Journal, 79(315): 573—578.
Banerjee, A.V., and A.F. Newman. 1998. “Information, the Dual Economy, and Development.” Review of Economic Studies, 65(4): 631—653.
Barry, J.M. 1998. Rising Tide: The Great Mississippi Flood of 1927 and How it Changed
America. Simon and Schuster.
Basu, S., and D.N. Weil. 1998. “Appropriate Technology and Growth.” Quarterly Journal
of Economics, 113(4): 1025—1054.
Besley, T., T. Persson, and D.M. Sturm. 2010. “Political Competition, Policy and
Growth: Theory and Evidence from the US.” Review of Economic Studies, 77(4): 1329—
1352.
Bleakley, H. 2007. “Disease and Development: Evidence from Hookworm Eradication in
the American South.” Quarterly Journal of Economics, 122(1): 73—117.
Boustan, L.P., M. Kahn, and P.W. Rhode. 2012. “Moving to Higher Ground: Migration Response to Natural Disasters in the Early Twentieth Century.” American Economic
Review: Papers and Proceedings, 102(3): 238—244.
Brannen, C.O. 1924. Relation of land tenure to plantation organization. GPO.
Brenner, R. 1986. “The social basis of economic development.” Analytical Marxism, 23—53.
Caselli, F., and W.J. Coleman. 2001. “The US Structural Transformation and Regional
Convergence: A Reinterpretation.” Journal of Political Economy, 109(3): 584—616.
Census Bureau, U.S. 1927. United States Census of Agriculture 1925. GPO.
Chay, K., and K. Munshi. 2012. “Black Networks After Emancipation: Evidence from
Reconstruction and the Great Migration.” Mimeo.
23
Cobb, J.C. 1994. The Most Southern Place on Earth: The Mississippi Delta and the Roots
of Regional Identity. Oxford Press.
Conley, T.G. 1999. “GMM Estimation with Cross Sectional Dependence.” Journal of
Econometrics, 92(1): 1—45.
David, P.A. 1975. Technical Choice, Innovation, and Economic Growth: Essays on American and British Experience in the Nineteenth Century. Cambridge Press.
Davis, A., B.B. Gardner, and M.R. Gardner. 2009. Deep South: A Social Anthropological Study of Caste and Class. University of South Carolina Press.
Day, R.H. 1967. “The Economics of Technological Change and the Demise of the Sharecropper.” American Economic Review, 57(3): 427—449.
Depew, B., P. Fishback, and P. Rhode. 2012. “New Deal or No Deal in the Cotton
South: The Effect of the AAA on the Agriculture Labor Structure.” Mimeo.
Dong, Z. 1993. “From Postbellum Plantation to Modern Agribusiness: A History of the
Delta and Pine Land Company.” PhD dissertation, Purdue University.
Donohue, J., and J. Heckman. 1991. “Continuous versus Episodic Change: The Impact
of Civil Rights Policy on the Economic Status of Blacks.” Journal of Economic Literature,
29: 1603.
Ellenberg, G.B. 2007. Mule South to Tractor South: Mules, Machines, and the Transformation of the Cotton South. University of Alabama Press.
Fishback, Price V. 1989. “Debt Peonage in Postbellum Georgia.” Explorations in Economic History, 26(2): 219—236.
Fishback, P., W. Horrace, and S. Kantor. 2005. “Did New Deal Grant Programs
Stimulate Local Economies? A Study of Federal Grants and Retail Sales During the
Great Depression.” Journal of Economic History, 65(1): 36—71.
Foster, A.D., and M.R. Rosenzweig. 1995. “Learning by Doing and Learning from Others: Human Capital and Technical Change in Agriculture.” Journal of Political Economy,
103(6): 1176—1209.
Gardner, B.L. 2002. American agriculture in the twentieth century: How it flourished and
what it cost. Harvard Press.
Grove, W.A., and C. Heinicke. 2003. “Better Opportunities or Worse? The Demise of
Cotton Harvest Labor, 1949-1964.” Journal of Economic History, 63(3): 736—767.
Habakkuk, H.J. 1962. American and British Technology in the Nineteenth Century: the
Search for Labour-saving Inventions. Cambridge Press.
Haines, M.R. 2005. “Historical, Demographic, Economic, and Social Data: The United
States, 1790-2002.”
24
Hayami, Y., and V.W. Ruttan. 1985. Agricultural Development: An International Perspective. Johns Hopkins Press.
Heckman, J.J., and B.S. Payner. 1989. “Determining the Impact of Federal Antidiscrimination Policy on the Economic Status of Blacks: A Study of South Carolina.” American
Economic Review, 79(1): 138—177.
Higgs, R. 1973. “Race, Tenure, and Resource Allocation in Southern Agriculture, 1910.”
Journal of Economic History, 33(1): 149—169.
Holley, D. 2000. The Second Great Emancipation: The Mechanical Cotton Picker, Black
Migration, and How They Shaped the Modern South. University of Arkansas Press.
Hornbeck, R. 2010. “Barbed Wire: Property Rights and Agricultural Development.” Quarterly Journal of Economics, 125(2): 767—810.
Hurst, W.M. 1933. Power and Machinery in Agriculture. USDA.
Kirby, J.T. 1987. Rural Worlds Lost: The American South, 1920-1960. LSU Press.
Kuznets, S. 1955. “Economic Growth and Income Inequality.” American Economic Review,
45(1): 1—28.
Langston, E.L., and B.H. Thibodeaux. 1939. “Plantation Organization and Operation
in the Yazoo-Mississippi Delta Area.” Mississippi Ag Bulletin, 682.
Lewis, W.A. 1954. “Economic Development with Unlimited Supplies of Labour.”
Manchester School, 22(2): 139—191.
Mandle, J.R. 1978. The Roots of Black Poverty: the Southern Plantation Economy after
the Civil War. Duke Press.
Margo, R.A. 1990. Race and Schooling in the South, 1880-1950: An Economic History.
University of Chicago Press.
Margo, R.A. 1991. “Segregated Schools and the Mobility Hypothesis: A Model of Local
Government Discrimination.” Quarterly Journal of Economics, 106(1): 61—73.
Murphy, K.M., A. Shleifer, and R.W. Vishny. 1989. “Industrialization and the Big
Push.” Journal of Political Economy, 97(5): 1003—1026.
Myrdal, G. 1944. An American Dilemma: the Negro Problem and Modern Democracy.
Transaction Publishers.
Naidu, S. 2010. “Recruitment Restrictions and Labor Markets: Evidence from the Postbellum U.S. South.” Journal of Labor Economics, 28(2): 413—445.
Naidu, S. 2012. “Suffrage, Schooling, and Sorting in the Post-Bellum U.S. South.” NBER
Working Paper 18129.
25
Olmstead, A.L., and P.W. Rhode. 2001. “Reshaping the landscape: the impact and
diffusion of the tractor in American agriculture, 1910—1960.” Journal of Economic History,
61(03): 663—698.
Parker, D.J. 2000. Floods. Vol. 1, Taylor & Francis.
Peterson, W., and Y. Kislev. 1986. “The Cotton Harvester in Retrospect: Labor Displacement or Replacement?” Journal of Economic History, 46(1): 199—216.
Ransom, R.L., and R. Sutch. 2001. One Kind of Freedom: The Economic Consequences
of Emancipation. Cambridge Press.
Raper, A. 1946. “The Role of Agricultural Technology in Southern Social Change.” Social
Forces, 25(1): 21—30.
Red Cross. 1928. “Reports and Statistics.” RG 200, Boxes 733-735, Folders 224.02, 224.031,
and 224.08.
Romer, P.M. 1986. “Increasing Returns and Long-run Growth.” Journal of Political Economy, 94(5): 1002—1037.
Rosengarten, T. 1975. All God’s Dangers: The Life of Nate Shaw. Knopf: New York.
Spencer, R. 1994. “Contested Terrain: The Mississippi Flood of 1927 and the Struggle to
Control Black Labor.” Journal of Negro History, 79(2): 170—181.
Steckel, R.H., and W.J. White. 2012. “Engines of Growth: Farm Tractors and
Twentieth-Century U.S. Economic Welfare.” NBER Working Paper 17879.
Tolnay, S.E., and E.M. Beck. 1995. A Festival of Violence: An Analysis of Southern
Lynchings, 1882-1930. University of Illinois Press.
Vaiden, M.G., J.O. Smith, and W.E. Ayres. 1931. “Making Cotton Cheaper, Can
Present Production Cost Be Reduced?” Mississippi Agricultural Bulletin, 290: 413—445.
Whatley, W.C. 1983. “Labor for the Picking: The New Deal in the South.” Journal of
Economic History, 43(4): 905—929.
Whatley, W.C. 1985. “A History of Mechanization in the Cotton South: The Institutional
Hypothesis.” Quarterly Journal of Economics, 100(4): 1191—1215.
Whatley, W.C. 1987. “Southern Agrarian Labor Contracts as Impediments to Cotton
Mechanization.” Journal of Economic History, 47(1): 45—70.
White, G.F. 1945. Human adjustment to floods: a geographical approach to the flood problem in the United States. University of Chicago Press.
Woodruff, N.E. 2003. American Congo: the African American freedom struggle in the
Delta. Harvard Press.
Wright, G. 1986. Old South, New South: Revolutions in the Southern Economy since the
Civil War. Basic Books: New York.
26
Figure 1. 1927 Flooded Region and Sample Counties (1900 Boundaries)
Notes: The 163 sample counties' boundaries are based on county definitions in 1900. County-level data are
adjusted to hold these boundaries fixed through 1970. The sample region flooded in 1927 is shaded gray, based on a
map compiled and printed by the US Coast and Geodetic Survey. The non-sample region is cross-hashed.
Excluded counties are missing outcome data in one of the analyzed years, have less than 15% of reported cropland
in cotton in 1920, or have a black population less than 10% of the total population in 1920.
27 9
9.9
9.1
10
9.2
10.1
9.3
10.2
9.4
10.3
Figure 2. Aggregate Changes in the Sample Region (AR, LA, MS, TN)
A. Log Black Population
B. Log Population
1900
1910
1920
1930
1940
1950
1960
1970
1900
1910
1920
1930
Year
1940
1950
1960
1970
1960
1970
1960
1970
Year
C. Log Value of Agricultural Capital
7
12
7.5
13
8
14
8.5
15
9
16
D. Log Number of Mules and Horses
1900
1910
1920
1930
1940
1950
1960
1970
1900
1910
1920
1930
Year
1940
1950
Year
E. Log Average Farm Size
2
4
3
4.5
4
5
5
6
5.5
F. Log Land Value per Farm Acre
1900
1910
1920
1930
1940
1950
1960
1970
1900
Year
1910
1920
1930
1940
1950
Year
Notes: Panels A-F report aggregated outcomes for the 163 sample counties in each year (Figure 1). Data are from
the US Census of Agriculture and the US Census of Population.
28 Figure 3. Estimated Differences in Black Population and Farm Capital, by Flood Share
Panel A. Log Black Population Share, Relative to 1920
Panel B. Log Value of Farm Equipment and Machinery, Relative to 1925
Notes: Panels A and B graph the estimated coefficients from column 1 of Table 2 and Table 4 (see the table notes
for details). The dashed lines indicate 95% confidence intervals.
29 Table 1. Baseline County Characteristics, by 1927 Flood Share
Log Difference by 1927 Flood Share:
Pre-Flood Levels:
Pre-Flood Changes:
Pre-Flood
Sample Mean Within-State
Controls
Within-State
Controls
(1)
(2)
(3)
(4)
(5)
Panel A. Population in 1920
Black Population Share
Black Population,
per 100 county acres
Population,
per 100 county acres
Panel B. Agriculture in 1925
Value of Farm Equipment,
per 100 county acres
Number of Mules & Horses,
per 100 county acres
Number of Tractors
per 100 county acres
Average Farm Size
Farmland Acres,
per 100 county acres
Value of Farm Land & Buildings,
per 100 farm acres
Value of Farm Land & Buildings,
per 100 county acres
0.461
[0.201]
2.99
[2.46]
6.24
[4.33]
0.782**
(0.101)
1.003**
(0.171)
0.220
(0.133)
0.449**
(0.133)
0.526*
(0.211)
0.077
(0.176)
-0.003
(0.022)
0.033
(0.064)
0.037
(0.057)
-0.055
(0.029)
-0.052
(0.082)
0.003
(0.071)
95.0
[60.9]
1.56
[0.84]
0.008
[0.010]
66.9
[21.4]
47.4
[17.3]
1606
[1316]
3370
[2094]
0.554**
(0.139)
0.422**
(0.141)
1.139**
(0.284)
-0.618**
(0.094)
-0.144
(0.102)
1.018**
(0.124)
0.875**
(0.168)
0.250
(0.178)
0.080
(0.172)
0.479
(0.390)
-0.417**
(0.101)
-0.244
(0.127)
0.702**
(0.162)
0.459*
(0.197)
-0.129
(0.079)
-0.080*
(0.040)
0.044
(0.115)
-0.048
(0.057)
0.017
(0.050)
-0.077
(0.045)
-0.272**
(0.046)
-0.350**
(0.061)
-0.076
(0.065)
-0.135*
(0.060)
-0.065
(0.060)
-0.200*
(0.081)
Number of Counties
163
163
163
163
163
Notes: Column (1) reports average baseline county characteristics in 1920 (Panel A) and 1925 (Panel B). All
variables are reported in levels (not logs) and the standard deviation is reported in parentheses. Column (2)
reports the within-state difference for each county characteristic (in logs) by the fraction of the county flooded in
1927: the coefficients are estimated by regressing the indicated county characteristic on the fraction of the
county flooded in 1927 and a state fixed effect, weighting by county size. Column (3) reports the estimated
difference when controlling also for each county's distance to the Mississippi river, geographic suitability for
cotton and corn, terrain ruggedness, and longitude and latitude. Column (4) reports the within-state difference in
pre-trends for each county characteristic (in logs): Panel A reports the change from 1910 to 1920 and Panel B
reports the change from 1920 to 1925. The coefficients are estimated by regressing the change in the indicated
county characteristic on the fraction of the county flooded in 1927 and a state fixed effect, weighting by county
size. Column (5) reports the estimated difference in pre-trends when controlling also for the above six countylevel variables. Robust standard errors are reported in parentheses: ** denotes statistical significance at 1%, *
denotes statistical significance at 5%.
30
Table 2. Estimated Differences in Population by Flood Share, Relative to 1920
Log Fraction Black
Decade:
1900
1910
1920
1930
1940
1950
1960
1970
(1)
0.051
(0.051)
0.003
(0.023)
0
-0.151**
(0.029)
-0.138**
(0.040)
-0.191**
(0.052)
-0.199**
(0.061)
-0.162*
(0.073)
(2)
0
-0.133**
(0.028)
-0.167**
(0.040)
-0.193**
(0.066)
-0.123
(0.079)
-0.110
(0.093)
Log Black Population
(3)
0.063
(0.116)
-0.033
(0.068)
0
-0.137**
(0.050)
-0.052
(0.066)
-0.117
(0.086)
-0.160
(0.105)
-0.310**
(0.116)
(4)
-
Log Population
(6)
-
0
(5)
0.011
(0.098)
-0.037
(0.062)
0
-0.137**
(0.045)
-0.075
(0.059)
-0.153
(0.083)
-0.189
(0.108)
-0.307*
(0.131)
0.011
(0.047)
0.086
(0.053)
0.074
(0.078)
0.039
(0.110)
-0.148
(0.131)
-0.018
(0.054)
0.044
(0.065)
0.045
(0.096)
0.003
(0.133)
-0.045
(0.153)
-
0
Counties
163
163
163
163
163
163
Notes: Each column reports estimated changes in the indicated outcome variable: changes in flooded counties
relative to changes in non-flooded counties, relative to the omitted year of 1920. Columns (1), (3), and (5)
report coefficients from regressing the outcome variable on the fraction of the county flooded in 1927, state-byyear fixed effects, and county fixed effects. Columns (2), (4), and (6) also control for county outcome values in
1900, 1910, and 1920, interacted with each year. All regressions are weighted by county size. Robust standard
errors clustered by county are reported in parentheses: ** denotes statistical significance at the 1% level, * at the
5% level.
31
Table 3. Estimated Differences in 1920-1930 Migration Rates by Flood Share
Panel A. All Matched People
Mean in Non-Flooded Counties
Difference in Flooded Counties
Counties with Matched People
Panel B. Black Population Only
Mean in Non-Flooded Counties
Difference in Flooded Counties
Counties with Matched People
Panel C. White Population Only
Mean in Non-Flooded Counties
Difference in Flooded Counties
Fraction Moving
Out-of-County
(1)
Fraction Moving
Out-of-State
(2)
Fraction Moving
Out-of-South
(3)
0.661
[0.130]
0.287
[0.136]
0.161
[0.123]
0.118**
(0.031)
0.113**
(0.033)
-0.003
(0.040)
162
162
162
0.698
[0.206]
0.245
[0.187]
0.136
[0.158]
0.139**
(0.046)
0.177**
(0.041)
0.068*
(0.030)
153
153
153
0.629
[0.152]
0.296
[0.183]
0.166
[0.162]
-0.069
(0.058)
0.034
(0.060)
-0.032
(0.050)
Counties with Matched People
156
156
156
Notes: Column 1 reports the fraction of people, matched between the 1920 census and 1930 census, that
appear in a different county. Column 2 reports the fraction of matched people that have left their state from
1920 to 1930, and Column 3 reports the fraction that have left the South from 1920 to 1930. Panel A
includes all matched people, panel B limits the sample to those people known to be black, and panel C limits
the sample to those people known to be white. Each panel and column reports the mean value in nonflooded counties and the standard deviation in brackets.
Each panel and column reports the difference in migration rate for flooded counties, relative to nonflooded counties, controlling for state fixed effects. All regressions are weighted by the fraction of matched
people in each county. Robust standard errors are reported in parentheses: ** denotes statistical
significance at the 1% level, * at the 5% level.
32
Table 4. Estimated Differences in Capital Intensity by Flood Share, Relative to 1925
Decade:
1900
1910
1920
1925
1930
Log Farm Capital
(1)
(2)
0.105
(0.161)
0.092
(0.142)
0.129
(0.085)
0
0
Log Mules & Horses
(3)
(4)
0.058
(0.099)
0.031
(0.080)
0.080
(0.043)
0
0
0.093
(0.086)
0.073
(0.079)
0.657**
(0.085)
0.594**
(0.090)
0.153**
(0.051)
0.167**
(0.052)
0.181*
(0.072)
0.130**
(0.049)
0.150**
(0.050)
0.182**
(0.067)
-0.283
(0.152)
-0.663**
(0.161)
-0.250
(0.135)
-0.610**
(0.139)
1935
1940
1945
Log Tractors
(5)
(6)
0
0
0.243
(0.207)
0.629**
(0.145)
0.954**
(0.268)
0.575*
(0.239)
1.411**
(0.229)
1.097**
(0.185)
0.188
(0.270)
0.846**
(0.189)
-0.003
(0.284)
0.711**
(0.177)
1950
1954
1960
1964
1970
1.096**
(0.148)
1.104**
(0.146)
Log Avg Farm Size
(7)
(8)
-0.081
(0.081)
0.047
(0.073)
-0.017
(0.052)
0
0
0.060
(0.051)
0.288**
(0.060)
0.264**
(0.069)
0.409**
(0.075)
0.566**
(0.085)
0.704**
(0.095)
1.148**
(0.132)
1.565**
(0.154)
1.582**
(0.160)
-0.013
(0.050)
0.078
(0.061)
0.026
(0.074)
0.136
(0.077)
0.254**
(0.092)
0.342**
(0.109)
0.498**
(0.141)
0.733**
(0.153)
0.581**
(0.151)
Counties
163
163
163
163
162
162
163
163
Notes: Each column reports estimated changes in the indicated outcome variable: changes in flooded counties
relative to changes in non-flooded counties, relative to the omitted year of 1920. Columns (1), (3), (5), and (7)
report coefficients from regressing the outcome variable on the fraction of the county flooded in 1927, state-byyear fixed effects, and county fixed effects. Columns (2), (4), (6), and (8) also control for county outcome
values in 1900, 1910, 1920, and 1925 (when available), interacted with each year. All regressions are weighted
by county size. Robust standard errors clustered by county are reported in parentheses: ** denotes statistical
significance at the 1% level, * at the 5% level.
33
Table 5. Estimated Differences in Farmland by Flood Share, Relative to 1925
Decade:
1900
1910
1920
1925
1930
1935
1940
1945
1950
1954
1960
1964
1970
Log Farmland
(1)
(2)
-0.176*
(0.070)
0.021
(0.063)
0.077
(0.047)
0
0
Log Value of Farmland
per farm acre
(3)
(4)
0.310**
(0.094)
0.019
(0.086)
0.272**
(0.048)
0
0
Log Value of Farmland
per county acre
(5)
(6)
0.134
(0.126)
0.040
(0.110)
0.350**
(0.064)
0
0
0.152**
(0.039)
0.239**
(0.048)
0.372**
(0.048)
0.507**
(0.067)
0.561**
(0.072)
0.632**
(0.082)
0.804**
(0.093)
0.925**
(0.103)
1.244**
(0.125)
-0.154**
(0.049)
-0.210**
(0.058)
-0.149**
(0.046)
-0.299**
(0.055)
-0.488**
(0.075)
-0.578**
(0.078)
-0.630**
(0.085)
-0.438**
(0.082)
-0.574**
(0.082)
-0.002
(0.051)
0.029
(0.073)
0.223**
(0.063)
0.208**
(0.068)
0.072
(0.098)
0.053
(0.112)
0.175
(0.126)
0.488**
(0.132)
0.670**
(0.145)
0.071
(0.042)
0.145**
(0.052)
0.277**
(0.059)
0.388**
(0.074)
0.451**
(0.084)
0.513**
(0.094)
0.651**
(0.113)
0.779**
(0.123)
1.079**
(0.154)
0.012
(0.052)
-0.007
(0.061)
-0.031
(0.057)
-0.154*
(0.074)
-0.143
(0.074)
-0.143
(0.073)
-0.159
(0.093)
-0.003
(0.083)
-0.075
(0.070)
-0.026
(0.054)
0.034
(0.079)
0.174*
(0.072)
0.247**
(0.083)
0.231*
(0.099)
0.288**
(0.108)
0.375**
(0.133)
0.646**
(0.136)
0.755**
(0.152)
Counties
163
163
163
163
163
163
Notes: Each column reports estimated changes in the indicated outcome variable: changes in flooded counties
relative to changes in non-flooded counties, relative to the omitted year of 1925. Columns (1), (3), and (5)
report coefficients from regressing the outcome variable on the fraction of the county flooded in 1927, state-byyear fixed effects, and county fixed effects. Columns (2), (4), and (6) also control for county outcome values in
1900, 1910, 1920, and 1925, interacted with each year. All regressions are weighted by county size. Robust
standard errors clustered by county are reported in parentheses: ** denotes statistical significance at the 1%
level, * at the 5% level.
34
Table 6. Estimated Differences in Black Population Share and Farm Capital: Robustness to Alternative Specifications
Baseline
Distance to
Estimates
MS River
Decade:
(1)
(2)
Panel A. Black Population Share
1930
-0.133**
-0.139**
(0.028)
(0.039)
1940
-0.167**
-0.150**
(0.040)
(0.045)
-0.193**
-0.185**
1950
(0.066)
(0.063)
-0.123
-0.152*
1960
(0.079)
(0.073)
1970
Controlling for Year-Interacted:
Suitability for
Terrain
Longitude &
Cotton & Corn
Ruggedness
Latitude
(3)
(4)
(5)
Controls in
(2) - (5)
(6)
-0.127**
(0.031)
-0.157**
(0.038)
-0.196**
(0.058)
-0.142**
(0.031)
-0.165**
(0.042)
-0.184**
(0.067)
-0.132**
(0.029)
-0.153**
(0.040)
-0.174**
(0.061)
-0.144**
(0.041)
-0.165**
(0.046)
-0.202**
(0.063)
-0.126
(0.069)
-0.122
(0.081)
-0.124
(0.081)
-0.104
(0.097)
-0.107
(0.077)
-0.081
(0.092)
-0.170*
(0.078)
-0.146
(0.094)
Treatment:
Population
Flooded
(7)
Propensity
Score
(8)
-0.106**
(0.031)
-0.149**
(0.039)
-0.181**
(0.066)
-0.128
(0.074)
-0.117
(0.087)
-0.120**
(0.033)
-0.147**
(0.045)
-0.166*
(0.078)
-0.086
(0.089)
-0.068
(0.103)
-0.110
-0.131
(0.093)
(0.088)
Panel B. Log Value of Farm Capital
1930
0.073
0.005
0.058
-0.033
0.028
-0.070
0.066
0.088
(0.079)
(0.098)
(0.083)
(0.089)
(0.080)
(0.104)
(0.087)
(0.085)
1940
0.594**
0.463**
0.555**
0.452**
0.521**
0.378**
0.629**
0.580**
(0.090)
(0.107)
(0.086)
(0.091)
(0.089)
(0.109)
(0.089)
(0.105)
1.027**
0.998**
0.905**
0.807**
1970
1.104**
0.818**
1.203**
1.124**
(0.150)
(0.168)
(0.120)
(0.153)
(0.146)
(0.168)
(0.188)
(0.164)
Counties
163
163
163
163
163
163
163
157
Notes: Column 1 reports baseline estimates from Table 2 column 2 (panel A) and Table 4 column 2 (panel B): changes in flooded counties relative to
changes in non-flooded counties, relative to the omitted year of 1920. The indicated outcome variable is regressed on the fraction of the county flooded in
1927, state-by-year fixed effects, county fixed effects, and year-interacted county outcome values in 1900, 1910, 1920, and 1925.
Columns 2 - 8 modify the baseline specification. Column 2 controls for counties' distance to the Mississippi river, interacted with each year. Column 3
controls for counties' suitability for cotton and corn, separately interacted with each year. Column 4 controls for counties' ruggedness, interacted with each
year. Column 5 controls for counties' longitude and lattitude, separately interacted with each year. Column 6 includes all of the controls from columns 2 - 5.
Column 7 instead measures flood intensity using the fraction of population affected by flooding in each county, as reported by the Red Cross. Column 8
controls for counties' estimated flood propensity score, interacted with each year, and limits the sample to counties with overlapping scores (estimated on a
county's black population share and fraction of cropland allocated to cotton in 1920).
All regressions are weighted by county size. Robust standard errors clustered by county are reported in parentheses: ** denotes statistical significance at
the 1% level, * at the 5% level.
35
Table 7. Estimated Differences in Black Population and Population: Robustness to Alternative Specifications
Baseline
Distance to
Estimates
MS River
Decade:
(1)
(2)
Panel A. Log Black Population
1930
-0.137**
-0.139*
(0.045)
(0.070)
1940
-0.075
-0.060
(0.059)
(0.074)
-0.153
-0.194
1950
(0.083)
(0.106)
-0.189
-0.286*
1960
(0.108)
(0.135)
1970
Controlling for Year-Interacted:
Suitability for
Terrain
Longitude &
Cotton & Corn
Ruggedness
Latitude
(3)
(4)
(5)
Controls in
(2) - (5)
(6)
-0.122*
(0.053)
-0.033
(0.069)
-0.111
(0.092)
-0.172**
(0.054)
-0.125
(0.067)
-0.181
(0.098)
-0.167**
(0.047)
-0.107
(0.059)
-0.213*
(0.084)
-0.170*
(0.076)
-0.107
(0.083)
-0.218
(0.115)
-0.141
(0.114)
-0.273*
(0.128)
-0.199
(0.126)
-0.278
(0.151)
-0.272*
(0.106)
-0.380**
(0.129)
-0.277
(0.141)
-0.344*
(0.165)
Treatment:
Population
Flooded
(7)
Propensity
Score
(8)
-0.168**
(0.044)
-0.111
(0.063)
-0.227**
(0.087)
-0.302**
(0.111)
-0.432**
(0.136)
-0.143**
(0.049)
-0.071
(0.064)
-0.092
(0.087)
-0.089
(0.112)
-0.146
(0.136)
-0.307*
-0.385*
(0.131)
(0.164)
Panel B. Log Population
1930
-0.018
-0.015
-0.011
-0.029
-0.030
-0.024
-0.062
-0.022
(0.054)
(0.066)
(0.062)
(0.058)
(0.050)
(0.072)
(0.056)
(0.061)
1940
0.044
0.038
0.061
0.026
0.026
0.029
0.012
0.055
(0.065)
(0.075)
(0.070)
(0.069)
(0.062)
(0.078)
(0.070)
(0.073)
1950
0.045
-0.038
0.051
0.027
-0.001
-0.042
-0.009
0.070
(0.096)
(0.106)
(0.099)
(0.104)
(0.091)
(0.110)
(0.095)
(0.103)
1960
0.003
-0.118
0.006
-0.018
-0.065
-0.112
-0.070
0.020
(0.133)
(0.148)
(0.131)
(0.145)
(0.131)
(0.149)
(0.131)
(0.138)
-0.063
-0.067
-0.126
-0.186
1970
-0.045
-0.192
-0.129
-0.049
(0.148)
(0.167)
(0.152)
(0.176)
(0.153)
(0.180)
(0.148)
(0.153)
Counties
163
163
163
163
163
163
163
157
Notes: Column 1 reports baseline estimates from Table 2 column 4 (panel A) and column 6 (panel B): changes in flooded counties relative to changes in
non-flooded counties, relative to the omitted year of 1920. The indicated outcome variable is regressed on the fraction of the county flooded in 1927, stateby-year fixed effects, county fixed effects, and year-interacted county outcome values in 1900, 1910, 1920, and 1925 (when available).
Columns 2 - 8 modify the baseline specification, as described in notes to Table 6. All regressions are weighted by county size. Robust standard errors
clustered by county are reported in parentheses: ** denotes statistical significance at the 1% level, * at the 5% level.
36
Table 8. Estimated Differences in Mules and Horses, Tractors, and Farm Size: Robustness to Alternative Specifications
Controlling for Year-Interacted:
Treatment:
Baseline
Distance to
Suitability for
Terrain
Longitude &
Controls in
Population
Propensity
Estimates
MS River
Cotton & Corn
Ruggedness
Latitude
(2) - (5)
Flooded
Score
Decade:
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
Panel A. Log Number of Mules and Horses
1930
0.130**
0.092
0.128**
0.080
0.106*
0.048
0.110
0.080
(0.049)
(0.055)
(0.047)
(0.054)
(0.047)
(0.057)
(0.058)
(0.057)
1940
0.182**
0.185*
0.200**
0.126
0.148*
0.155*
0.179*
0.224**
(0.067)
(0.078)
(0.068)
(0.068)
(0.065)
(0.075)
(0.089)
(0.075)
-0.250
-0.364*
-0.199
-0.218
-0.328**
-0.242
1954
-0.252
-0.164
(0.135)
(0.159)
(0.138)
(0.140)
(0.111)
(0.127)
(0.155)
(0.151)
-0.610**
-0.672**
1960
-0.553**
-0.469**
-0.675**
-0.460**
-0.590**
-0.507**
(0.139)
(0.166)
(0.142)
(0.143)
(0.118)
(0.135)
(0.146)
(0.158)
Panel B. Log Number of Tractors
1930
0.629**
0.589**
0.571**
0.566**
0.597**
0.473*
0.666**
0.396*
(0.145)
(0.182)
(0.158)
(0.175)
(0.146)
(0.193)
(0.169)
(0.188)
1940
1.411**
1.254**
1.177**
1.184**
1.372**
0.951**
1.451**
0.776**
(0.229)
(0.282)
(0.213)
(0.259)
(0.208)
(0.261)
(0.259)
(0.252)
1954
0.846**
0.596*
0.712**
0.607**
0.783**
0.403
0.838**
0.694**
(0.189)
(0.230)
(0.175)
(0.210)
(0.189)
(0.209)
(0.240)
(0.220)
0.553**
0.595**
0.659**
0.455*
1970
0.711**
0.574**
0.722**
0.600**
(0.169)
(0.197)
(0.177)
(0.204)
(0.177)
(0.220)
(0.226)
(0.209)
Panel C. Log Average Farm Size
1930
-0.013
0.015
-0.025
0.012
0.025
0.037
0.031
-0.013
(0.050)
(0.058)
(0.050)
(0.054)
(0.053)
(0.058)
(0.054)
(0.046)
1940
0.026
0.130
0.087
0.024
0.116
0.185*
0.071
-0.033
(0.074)
(0.076)
(0.073)
(0.079)
(0.074)
(0.082)
(0.083)
(0.072)
1954
0.465**
0.354**
0.473**
0.609**
0.342**
0.463**
0.345**
0.226*
(0.097)
(0.116)
(0.097)
(0.106)
(0.109)
(0.108)
(0.114)
(0.105)
1970
0.775**
0.521**
0.760**
0.723**
0.581**
0.532**
0.540**
0.374**
(0.139)
(0.160)
(0.140)
(0.160)
(0.151)
(0.167)
(0.165)
(0.139)
Counties
163
163
163
163
163
163
163
157
Notes: Column 1 reports baseline estimates from Table 4 column 4 (panel A), column 6 (panel B), and column 8 (panel C). Columns 2 - 8 modify the
baseline specification, as described in notes to Table 6. Panel B includes 162 or 156 counties. All regressions are weighted by county size. Robust standard
errors clustered by county are reported in parentheses: ** denotes statistical significance at the 1% level, * at the 5% level.
37
Table 9. Estimated Differences in Farmland and Value of Farmland: Robustness to Alternative Specifications
Controlling for Year-Interacted:
Treatment:
Baseline
Distance to
Suitability for
Terrain
Longitude &
Controls in
Population
Propensity
Estimates
MS River
Cotton & Corn
Ruggedness
Latitude
(2) - (5)
Flooded
Score
Decade:
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
Panel A. Log Farmland
1930
0.071
0.022
0.060
0.014
0.050
-0.023
0.093*
0.052
(0.042)
(0.047)
(0.042)
(0.047)
(0.039)
(0.046)
(0.047)
(0.049)
1940
0.277**
0.251**
0.276**
0.216**
0.242**
0.203**
0.315**
0.256**
(0.059)
(0.062)
(0.053)
(0.064)
(0.055)
(0.058)
(0.070)
(0.063)
1954
0.499**
0.460**
0.437**
0.408**
0.513**
0.438**
0.540**
0.471**
(0.095)
(0.104)
(0.079)
(0.100)
(0.094)
(0.114)
(0.126)
(0.104)
1970
1.026**
1.043**
1.002**
0.943**
1.079**
0.946**
1.185**
0.947**
(0.152)
(0.176)
(0.141)
(0.177)
(0.154)
(0.198)
(0.218)
(0.183)
Panel B. Log Value of Agricultural Land and Buildings, per Farm Acre
1930
0.012
-0.023
0.011
0.001
-0.029
-0.043
-0.027
0.004
(0.052)
(0.058)
(0.053)
(0.057)
(0.057)
(0.064)
(0.049)
(0.051)
1940
-0.031
-0.089
-0.016
-0.062
-0.052
-0.110*
-0.050
-0.021
(0.057)
(0.057)
(0.051)
(0.058)
(0.057)
(0.055)
(0.058)
(0.062)
1954
-0.180*
-0.169*
-0.199**
-0.351**
-0.143
-0.302**
-0.190**
-0.128
(0.072)
(0.078)
(0.074)
(0.077)
(0.073)
(0.075)
(0.070)
(0.081)
1970
-0.105
-0.119
-0.172*
-0.301**
-0.075
-0.231**
-0.096
-0.057
(0.072)
(0.074)
(0.073)
(0.080)
(0.070)
(0.081)
(0.069)
(0.074)
Panel C. Log Value of Agricultural Land and Buildings, per County Acre
1930
-0.026
-0.090
-0.023
-0.059
-0.082
-0.119
-0.043
-0.046
(0.054)
(0.061)
(0.052)
(0.060)
(0.051)
(0.065)
(0.056)
(0.061)
1940
0.174*
0.085
0.207**
0.115
0.109
0.055
0.183*
0.154
(0.072)
(0.073)
(0.064)
(0.077)
(0.066)
(0.073)
(0.082)
(0.081)
1954
0.274*
0.228*
0.065
-0.042
0.288**
0.015
0.267*
0.280*
(0.111)
(0.115)
(0.091)
(0.100)
(0.108)
(0.119)
(0.127)
(0.118)
1970
0.759**
0.681**
0.483**
0.401**
0.755**
0.452*
0.810**
0.752**
(0.155)
(0.166)
(0.129)
(0.152)
(0.152)
(0.183)
(0.198)
(0.180)
Counties
163
163
163
163
163
163
163
157
Notes: Column 1 reports baseline estimates from Table 5 column 2 (panel A), column 4 (panel B), and column 6 (panel C). Columns 2 - 8 modify the
baseline specification, as described in notes to Table 6. All regressions are weighted by county size. Robust standard errors clustered by county are reported
in parentheses: ** denotes statistical significance at the 1% level, * at the 5% level.
38
Table 10. Estimated Relative Changes in Counties within 50km of other Major Southern Rivers
Decade:
1930
Log
Fraction Black
(1)
0.009
(0.012)
Log Black
Population
(2)
0.018
(0.026)
Log
Population
(3)
0.008
(0.021)
Log Value of
Farm Capital
(4)
-0.045
(0.038)
0.025
(0.016)
0.025
(0.032)
0.001
(0.026)
-0.092
(0.057)
0.016
(0.023)
0.007
(0.043)
-0.016
(0.037)
1935
1940
Log Mules
& Horses
(5)
-0.012
(0.022)
-0.004
(0.023)
-0.035
(0.028)
1945
1950
1954
1960
-0.077
(0.059)
-0.010
(0.030)
-0.011
(0.054)
-0.011
(0.054)
-0.016
(0.037)
0.003
(0.069)
0.003
(0.069)
-0.001
(0.066)
1964
1970
-0.190*
(0.088)
Log Avg
Farm Size
(6)
0.052*
(0.023)
0.014
(0.030)
0.041
(0.031)
0.065
(0.037)
0.055
(0.038)
0.043
(0.039)
0.056
(0.043)
0.078
(0.047)
0.091
(0.049)
Log
Farmland
(7)
0.021
(0.016)
0.024
(0.018)
0.029
(0.020)
0.002
(0.022)
-0.008
(0.026)
-0.006
(0.031)
-0.011
(0.043)
-0.009
(0.047)
-0.031
(0.060)
Log Land Val Log Land Val
per farm acre per county acre
(8)
(9)
-0.013
0.001
(0.027)
(0.024)
0.002
-0.003
(0.036)
(0.031)
-0.015
-0.023
(0.037)
(0.034)
-0.042
-0.071
(0.041)
(0.039)
0.016
-0.031
(0.041)
(0.045)
-0.028
-0.084
(0.047)
(0.054)
-0.042
-0.095
(0.047)
(0.064)
-0.054
-0.105
(0.046)
(0.070)
-0.020
-0.078
(0.041)
(0.067)
Counties
243
243
243
243
243
243
243
243
243
Notes: Each column reports estimated changes in the indicated outcome variable: changes in counties within 50km of a major river relative to changes in
counties within 50km - 150km of a major river, relative to the omitted year of 1920 or 1925. The sample is restricted to Southern counties within 150km of a
major river, excluding all counties in the main sample region (Figure 1). The indicated outcome variable is regressed on a dummy for whether the county is
within 50km of a major river, state-by-year fixed effects, county fixed effects, and county outcome values in 1900, 1910, 1920, and 1925 (when available),
interacted with each year. All regressions are weighted by county size. Robust standard errors clustered by county are reported in parentheses: ** denotes
statistical significance at the 1% level, * at the 5% level.
39
Table 11. Estimated Changes in Counties Bordering the Flooded Region, Relative to Counties 100km Away
Decade:
1930
Log
Log Black
Log
Log Value of
Log Mules
Log
Log Avg
Log
Fraction Black
Population
Population
Farm Capital
& Horses
Tractors
Farm Size
Farmland
(1)
0.019
(0.023)
(2)
0.003
(0.048)
(3)
0.003
(0.035)
(4)
0.010
(0.064)
(6)
0.045
(0.140)
0.035
(0.034)
0.007
(0.057)
0.010
(0.044)
-0.048
(0.066)
(5)
-0.033
(0.046)
-0.045
(0.042)
-0.078
(0.042)
0.075
(0.052)
0.050
(0.081)
0.050
(0.074)
0.052
(0.067)
0.084
(0.117)
(7)
-0.022
(0.027)
-0.029
(0.030)
-0.060*
(0.030)
-0.035
(0.040)
-0.078
(0.053)
-0.112
(0.057)
0.046
(0.121)
-0.105
(0.064)
-0.070
(0.068)
-0.001
(0.064)
(8)
-0.015
(0.028)
-0.014
(0.037)
-0.032
(0.042)
-0.007
(0.053)
-0.048
(0.056)
-0.081
(0.061)
-0.130
(0.073)
-0.116
(0.082)
-0.111
(0.111)
1935
1940
1945
1950
1954
1960
0.101
(0.064)
0.068
(0.107)
0.064
(0.113)
0.102
(0.076)
0.077
(0.127)
0.094
(0.140)
0.260
(0.218)
0.179
(0.158)
0.077
(0.078)
1964
1970
-0.036
(0.084)
Log Land Val
Log Land Val
per farm acre per county acre
(9)
0.060
(0.039)
0.022
(0.050)
0.047
(0.046)
0.102*
(0.051)
0.129*
(0.056)
0.114*
(0.053)
0.076
(0.055)
0.151**
(0.053)
0.118**
(0.042)
(10)
0.048
(0.043)
0.004
(0.059)
0.018
(0.055)
0.106
(0.064)
0.093
(0.063)
0.035
(0.068)
-0.055
(0.062)
0.034
(0.074)
0.027
(0.099)
Counties
94
94
94
94
94
94
94
94
94
94
Notes: Each column reports estimated changes in the indicated outcome variable: changes in counties bordering the flooded region relative to changes in
counties 100km from the flooded region, relative to the omitted year of 1920 or 1925. The sample is restricted to the 94 main sample counties with no
flooding (Figure 1). The indicated outcome variable is regressed on the (negative) distance from the flooded region in 100km units, state-by-year fixed
effects, county fixed effects, and county outcome values in 1900, 1910, 1920, and 1925 (when available), interacted with each year. All regressions are
weighted by county size. Robust standard errors clustered by county are reported in parentheses: ** denotes statistical significance at the 1% level, * at the
5% level.
40
Appendix Figure 1. Delta Land and Pine Company Profits and Cotton Production
$150
300
$100
250
$50
200
150
$0
1911
1916
1921
1926
1931
1936
1941
1946
1951
1956
1961
-$50
100
-$100
50
-$150
0
Profits (2010 dollars)
Bales/Acre (10s)
Cotton Price (cents/pound)
Notes: Delta Land and Pine Company Profits and Bales per Acre are from Dong 1993. Cotton prices are from
Historical Statistics of the United States. On the left axis are profits, measured in 2010 dollars. On the right axis
are: cotton prices, measured in cents per pound; and cotton bales per acre, measured in units of 10.
41